224 IEEE TRANSACTIONS ON EDUCATION, VOL. 41, NO. 3, AUGUST 1998 Experimental Verification of the Physics and Structure of the Bipolar Junction Transistor I. M´ artil, Member, IEEE, J. M. Mart´ ın, S. Garc´ ıa, and G. Gonz´ alez-D´ ıaz Abstract— We present an electrical characterization of dis- crete Bipolar Junction Transistor (BJT) devices with nonuniform doped emitter and base zones. The measurement of the – and – characteristics of the emitter–base and the collector–base junctions and the common emitter current gain allows to deter- mine relevant parameters of the device. These are the built-in voltage of both junctions, the impurity gradient profiles, the electrical area of both junctions, the base and the emitter Gummel numbers, and the collector doping. The whole experiment can be conducted in a laboratory session of 3–4-hour length and it is specifically addressed to students taking lectures in semiconduc- tor device physics. The results obtained give a deep insight into both the physical structure and the physical processes involved in the transistor behavior. Index Terms—Bipolar junction transistor, device parameters, physics of BJT’s. I. INTRODUCTION B IPOLAR junction transistors (BJT), together with field- effect transistors (JFET, MOSFET, etc.) are the most widely known semiconductor devices. Students of semicon- ductor device physics devote several weeks to understand the complex physical processes involved in the transistor action. There are excellent books which provide a comprehensive treatment of the physics of both BJT and FET’s [1]–[3]. From the practical point of view, laboratory experiments with BJT’s are mainly concerned with two aspects: determi- nation of the equivalent circuit of the BJT [4] and simulation of the BJT behavior with proper programs, PSPICE being the most popular among them [5]. However, experiments with the BJT are scarce, and mainly focused on the practical applications of the device (i.e., signal amplification, phase shift, etc.). The electrical characteriza- tion of a BJT has been used to obtain information about semiconductor parameters (bandgap of Si and Ge) [6], with no emphasis in the determination of device parameters like doping levels, minority-carrier lifetimes, etc. Quite accurate information about the physical structure of the BJT can be deduced by means of the – characteristics of the device starting from the Ebers–Moll model [7] and introducing simple changes into the equations for a more suitable description of double diffused transistors [2]. On the other hand, by means of – measurements, it is possible to obtain ad- ditional data from the device. However, these experiments Manuscript received November 13, 1994; revised March 21, 1998. The authors are with the Departamento Electricidad y Electr´ onica, Fac. F´ ısicas, Universidad Complutense de Madrid, 28040 Madrid, Spain (e-mail: imartil@eucmax.sim.ucm.es). Publisher Item Identifier S 0018-9359(98)05719-7. can be found only in research papers published during the last thirty years (see, for instance, [1, Ch. 3] and references, therein). From an educational point of view, we have not found undergraduate-level laboratory experiments focused on the practical demonstrations of the main aspects of the BJT theory. In this paper, we propose a set of electrical measurements to be performed on discrete Si BJT’s to measure several key device parameters, which are responsible for the transistor ac- tion. The measurements are designed to be done in a 3–4-hour laboratory session and are focused on a second course on semiconductor device physics. II. THEORY Present-day discrete Si BJT devices are made using the same fabrication processes involved in integrated circuit fabrication. As a consequence, the structure of BJT commercial devices is more complex than the ideal structure: the doping of the base and the emitter zones is made by diffusion or ion implantation of impurities into the uniformly doped collector zone. This means that doping profiles of both the emitter and the base are not uniform and, therefore, it is necessary to use Gummel numbers to characterize their respective doping levels instead of impurity-concentration values [8]. Due to the resulting doping profiles of the three zones, the characteristics of both emitter–base and collector–base junctions (EBJ and CBJ) are dependent on the bias-voltage range of measurement, behaving as linearly graded junctions for small dc bias voltages (both of them) and as an abrupt junction for large reverse dc bias voltages (the CBJ) [2]. In Fig. 1 we present a scheme of the physical structure and the impurity doping profiles for a typical n-p-n BJT double-diffused discrete device [3]. The – characteristic of the EBJ, when the collector is short-circuited to the base (i.e., when the BJT works in the so- called transdiode regime where ) [6], can be written as follows: (1) where is the EBJ dc bias, is the EBJ area, and is the intrinsic Si carrier concentration. and are, respectively, the base and emitter Gummel numbers, defined as follows [3], [9]: (2) 0018–9359/98$10.00  1998 IEEE